Produced in partnership with BNY 

Much has been written about the benefits of a total portfolio approach (TPA), how it differs from traditional strategic asset allocation (SAA) models and the changes required to the task of portfolio construction. Under TPA, investment risk factors, rather than asset classes and their benchmarks, are the primary diversification lever for assessing opportunities for total fund contribution. The shift toward a total portfolio approach requires significant changes to the data and operations that support investment teams.

Governance models, including risk appetite, will shift. Culture, motivation, skills, structure and alignment to objectives must transform. There are greater demands in terms of agility and the ability to access market opportunities in real time. Support functions and service partnerships will also need to adapt, as funds re-evaluate their investment strategies.

TPA is not a prescriptive methodology or model; it is descriptive. Asset owners often apply TPA principles in diverse ways to meet their fund’s objectives and practical constraints. In the 2024 article The Rise of Total Portfolio Approach, the author says this succinctly: “There is no uniform checklist meant to represent an instruction booklet to assemble TPA at your organisation.”

In the superannuation industry, additional challenges require funds to continuously assess how data, technology and operations support their investment processes. TPA cannot ignore industry-accepted views that members, advisers and market regulators are all accustomed to. Member investment options are designed for varying risk-return appetites which are grounded in SAA-based asset class concepts, such as “growth” or “defensive”. What’s more, these investment options are often implemented via complex, cross-invested asset class structures designed more for implementation efficiency than strategic agility.

Furthermore, a fund’s in-house investment operations, functions and technology, as well as their administrator’s operations and technology, will have evolved over many years to report asset classes constructed of portfolios, not decision-making based on risk factors.

These are some of the practical considerations that factor into design and implementation, and which suggest that change is on the horizon.

Demand for next-generation data platforms

The learning experiences of Australia’s superannuation industry run deep and span a wide array of data management technologies. Some funds have made substantial progress on their data management journey, having grappled with the challenges of managing disparate data sources with varying quality, timeliness and accuracy, while delivering value to their business functions.

“TPA thinking” gives rise to new data management priorities:

· Integrated “book of record” capabilities. Because TPA demands agility, an investment book of record with position data fed by administrators may not be timely or represent transactions completely enough to support the daily calculation of your chosen investment risk factors. Data platforms with native integration of an investment book of record’s capabilities go a long way towards closing this gap on timeliness and completeness. These capabilities can now be accessed as a service or via outsourced middle-office arrangements.

· Private market data access and total portfolio view. The ability to unify a whole-of-fund total portfolio view across public and private assets supported by general partner (GP) data is critical to supporting many TPA needs. GP data across private equities, real assets and private debt markets can vary in terms of granularity, timeliness, quality and scope, to say nothing of collection, formatting and delivery effort. Because private market data is not homogenous, it is not surprising that funds are turning toward specialist private market data services, AI and intelligent document processing to advance this priority.

· Content enablement and semantic models. As a key principle of TPA thinking is to allocate capital to the best investment right now, data content requirements will never remain static. Investment teams want easy access to new data content, and they want it “use-case ready.” Data platforms that can easily enable new data content using inbuilt or open semantic models (ways to represent the data) will easily deliver a time-to-value advantage for investment teams.

· Look-through and unitisation lag. Many super funds implement their member investment options via a complex, cross-invested asset class structure which introduces demand to synthesize a top-down investment option look-through to the underlying asset holdings. However, this poses data challenges in an industry heavily dependent on the lagged T+1 accounting cycle needed to unitise an investment option’s asset allocation position and creates operational challenges for real-time liquidity tracking as funds seek opportunities to trade across time zones.

Investment analysts as citizen developers

The total portfolio approach relies on portfolio construction teams for data experimentation, deep analysis, innovation, intellectual property development and knowledge sharing. Today’s investment analysts demand access to diverse data sources, including risk, pricing, performance, security, issuer, exposures, climate, economic and geopolitical data. And many are adept at skilfully blending complex data sets using programming languages such as Python and R.

Investment analysts are evolving into “citizen developers,” creating essential TPA datasets such as risk, exposure and strategy targets from large datasets. These non-traditional developers combine their skillsets with data experimentation and investment theory, and their demands for quick access to new tools will continue to grow as AI becomes more pervasive and affordable.

Powering capability via AI

Running an investment data platform is operationally challenging and requires specialist data skills to support risk, portfolio construction and investment operations teams. Features such as data catalogues, lineage and data quality controls are in high demand and are expected to be AI-enabled to help organisations maintain the vast volume of metadata these functions rely on.

There is symbiosis at play. AI-powered workflows are driving efficiencies in metadata maintenance across large datasets, which is integral to supporting natural language agents that can help investment analysts rapidly respond to market conditions and generate new investment insights.

Furthermore, we anticipate that agentic workflows will become prominent by unlocking new insights. This will give analysts the ability to work with data at scale, effectively synthesizing new information from internal and external data sources across multiple business domains, while automating much of what is handled manually today and creating net capacity.  It shrinks the operational footprint and enables human curation of strategic insights that can drive investment performance.

In a 2024 Thinking Ahead Institute survey of 26 leading asset owners committed to a total portfolio approach, 84 per cent of survey respondents indicated that AI will either be an integral component of their technology infrastructure or they will develop projects that will lead to this.

Given this finding, it seems clear that data platforms with AI embedded foundations will play a critical role in delivering higher value investment insights via agentic workflows.

Data-enabled operating models

Operating model integration remains a critical success factor for any data platform initiative. Investment operations are no longer just agents sending instructions or moving cash, they are data stewards of vital records that drive data accuracy and timeliness across the public markets middle-office, private market placements, specialist sub-advisers, custody and fund accounting. Investment operations are core to determining what is true versus what is false.

As operating model design and data platform design continue to converge into a single design activity, we expect this focus to intensify as the superannuation segment thematically gravitates toward TPA principles.

A total portfolio approach offers transformative opportunities for the superannuation segment, emphasising investment risk factors over traditional asset classes. To harness the potential of TPA, funds must continuously evaluate and adapt their data and technology strategy, their AI capabilities and the role of operations to effectively support their evolving investment strategy.

Click here to learn more about BNY’s data and analytics solutions.